Ovarian cancer is the most deadly gynecologic cancer. The main reason for a high death rate is that this cancer is seldom detected until it is well advanced. And early detection is difficult because there is little known about how this deadly cancer begins and metastasizes. Although surgery and chemotherapy have gradually improved over the last 20 years, most women with ovarian cancer who successfully respond to initial chemotherapy relapse and eventually die. To reduce ovarian-cancer deaths, it is therefore essential to detect the cancer early and to more effectively treat advanced ovarian cancers. Early detection and effective treatment of ovarian cancer will require better understanding of the molecular mechanisms underlying deadly ovarian cancer. As part of my postdoctoral training, I have developed a genetically engineered mouse model of ovarian cancer by conditionally disabling the two genes: the DICER gene, encoding an essential enzyme for microRNA biosynthesis, and the PTEN gene, encoding a tumor suppressor. These mutant mice lacking both genes in the ovary and fallopian tube develop a highly aggressive metastatic serous epithelial cancer, which closely resembles human ovarian cancer. This NRSA application will therefore focus on characterizing and using these mice to further understand ovarian cancer in women. The general hypothesis of my proposal is that overall decrease of microRNA levels in a tumor-prone environment promotes development of ovarian cancer. Specific Aim1 will investigate the origin and early tumor process of epithelial ovarian cancer. To help identify early marker genes of ovarian cancer, gene-expression analyses will be performed on the ovarian tumor tissues sampled from these mutant mice. Specific Aim2 will define the metastatic nature of these mouse ovarian cancers that mirror human ovarian cancer. The metastatic process will be investigated in these mice also by gene-expression analyses. In addition, tumor cells from this mouse cancer will be cultured to study the molecular pathways of ovarian-cancer progression and metastasis. Together, these approaches will yield useful drug targets for treating advanced ovarian cancers. Specific Aim3 will involve creating additional mouse models of ovarian cancer. Because most patients with deadly ovarian cancers carry mutations in the p53 tumor-suppressor gene, a mutation in the p53 gene will be incorporated into the mutant mice that already lack the DICER and PTEN genes. This new mutant-mouse model would be predicted to develop ovarian cancer that is genetically more similar to human ovarian cancer. The primary goal of this NRSA proposal will be to define the origin and molecular pathways of highly metastatic epithelial ovarian cancer using genetically engineered mouse models.